For the marketer who owns one business's AI search presence.
1. Signing in
You received a personal link that looks like
schema-auditor.zigfeld.com/?token=…. Open it
once and the browser remembers you; from then on the plain address works.
The link is yours alone. If you ever see a permissions error after clicking
someone else's link, reopen your own link and you are back.
2. Your dashboard in 60 seconds
The dashboard opens on What to do next: a short list of the highest
value moves, each tagged markup (we generate it, you
paste it) or site content (your team writes it, we
draft the outline). Every item links straight to where the work happens.
Citation Score and the Engine Gap: one number for how AI
assistants treat you, and a per assistant breakdown. Download the stat card
image to share it internally.
What the gap costs: a calculator translates un-cited AI answers
into an estimated business cost. What it counts depends on your vertical.
A hotel sees it in direct-booking commission (enter rooms, rate,
occupancy); most other businesses see it in lost inquiries (enter a monthly
lead value). Your workspace shows the version built for you.
Content opportunities: the pages worth writing, derived from your
own data, with one-click brief drafting.
3. Audit: your site, scored and explained
The Audit stage crawls your site, scores it 0 to 100, and explains itself.
Nothing here ever touches your live website.
The verdict and the analyst's read lead: what the score
means, the weakest link in plain terms, and the single most valuable next
move. Computed from your scan, never generated prose.
What to do next merges every recommendation into one ranked list
with point values.
Customer questions shows which real questions an assistant can
answer from your markup, and exactly what would answer the misses.
Google Business Profile grades your map pack presence: profile
found, categorized, healthy reviews, and whether the website button sends
customers to your site or to a third party that intercepts the lead.
Google rank check shows where you actually rank for your own name
and your market's head terms, and names every third party outranking
you.
Titles & descriptions, rewritten appears after a rewrite run:
copy-paste title tags and meta descriptions for every flagged page.
Marketing / Technical toggle. The audit opens in
marketing view with the technical layers folded. The toggle at the top opens
everything for your web team, and the browser remembers your choice.
4. Benchmark: what AI says about you
The Benchmark stage measures the answers themselves.
Run a census to put a set of real customer questions to four
assistants. Each prompt shows a check when your own site earned the citation;
hover any source badge to see the exact page it cites.
Fact health checks whether assistants get your hours, phone,
pricing, and policies right, with the exact wrong quote when one is off.
The earned-media hit list names every page assistants cite instead
of you, with receipts, plus an authority chip:
green means the domain is within your reach to
outrank; amber means pitch to be featured on
it.
Competitors: discovery finds and verifies rivals in under a
minute, or add them by URL. Benchmarks score you side by side.
5. Opportunities and briefs: content that earns citations
Opportunity cards are content plays derived from your own data: customer
questions no one answers, and comparison pages assistants prefer to cite.
Press Draft brief and you get publish-ready copy grounded in your
verified facts.
The [CONFIRM] boxes matter. Anything only your team
can know appears as a highlighted fill-in. Complete them and press
Save confirmed facts; the brief becomes the publishable
version.
6. Fix, Publish, Verify: shipping the markup
Three stages, one loop, nothing automatic on your live site.
Fix is the checklist: every finding with its remedy, automatic
fixes applied to the proposal, point values on each item.
Publish hands you copy-paste packages for your CMS, gated so
nothing flagged ever ships. Pick your CMS and paste where it says.
Verify re-scans after you paste. It passes when the live site
scores at least what the proposal promised, so improvement is proven, not
assumed.
7. Watching work run
The queue chip in the header shows anything running: crawls,
censuses, briefs. Each job reports its phase and step so a ten minute census
never looks stuck. The small × removes a finished row, or cancels a
running job after a confirmation.
8. Feedback and what's new
The feedback button floats at the bottom right of every screen. It
captures your whole screen plus your notes and sends both to us, for bugs and
ideas alike. The What's New screen appears once after each release;
press OK and it stays away until the next one.
9. Glossary
Term
What it means
AI Search Readiness Score
0 to 100. How well your website's
structured markup lets AI assistants read, trust, and cite you. This is the
score on every dashboard card and audit.
Citation Score
0 to 100. How AI assistants actually treat you
today: are you mentioned, and do answers link your own site?
Visibility
The share of test questions where the assistant
mentioned your business at all.
Cited
The share of answers that linked your own website.
Mentioned but not cited means the lead goes elsewhere.
Census
One measurement run: a set of real customer questions
put to Claude, ChatGPT, Gemini, and Perplexity, with every answer stored as
evidence.
Engine Gap
The spread between assistants. A business can be a
Claude regular and a ChatGPT stranger; the gap tells you where the work
is.
DR (authority)
0 to 100 domain strength from referring links.
Shown on the sites that get cited instead of you: green means it is within
your reach to outrank, amber means pitch to be featured on it.
Markup / Site content
The two kinds of fixes. Markup changes
are generated for you and pasted into your CMS. Site content is a page your
team writes, usually from a drafted brief.
Brand Canon
Your verified facts (address, phone, services,
policies). Everything the platform writes is grounded in the canon and your
own pages; nothing is invented.